Upload folder using huggingface_hub
Browse files- config.json +2 -2
- model.safetensors +1 -1
- optimizer.pt +2 -2
- rng_state.pth +2 -2
- scheduler.pt +2 -2
- trainer_state.json +252 -728
- training_args.bin +2 -2
config.json
CHANGED
|
@@ -1,5 +1,4 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "facebook/dinov2-large",
|
| 3 |
"apply_layernorm": true,
|
| 4 |
"architectures": [
|
| 5 |
"Dinov2ForImageClassification"
|
|
@@ -64,6 +63,7 @@
|
|
| 64 |
"stage24"
|
| 65 |
],
|
| 66 |
"torch_dtype": "float32",
|
| 67 |
-
"transformers_version": "4.
|
|
|
|
| 68 |
"use_swiglu_ffn": false
|
| 69 |
}
|
|
|
|
| 1 |
{
|
|
|
|
| 2 |
"apply_layernorm": true,
|
| 3 |
"architectures": [
|
| 4 |
"Dinov2ForImageClassification"
|
|
|
|
| 63 |
"stage24"
|
| 64 |
],
|
| 65 |
"torch_dtype": "float32",
|
| 66 |
+
"transformers_version": "4.55.0",
|
| 67 |
+
"use_mask_token": true,
|
| 68 |
"use_swiglu_ffn": false
|
| 69 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1217542512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0008d0b179f0af6309b99b8c6fabf134a11ed6bdd66ded13444541a4a3e6fbcb
|
| 3 |
size 1217542512
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:48abcb40e932df8c553badeafb3b0a18b0f7d781ab7f8395c3cf20c85c189a66
|
| 3 |
+
size 2435342411
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4395a530497f268c762a08cd4fcee96c37463e64496c3f6e4a0c83bb1f5337f6
|
| 3 |
+
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2e6312b4ddd707deada4d923ee63434e3e31f0333703f16ac94f13036d755de
|
| 3 |
+
size 1465
|
trainer_state.json
CHANGED
|
@@ -1,858 +1,382 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
| 4 |
-
"
|
|
|
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss":
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss":
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss":
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch":
|
| 34 |
-
"
|
| 35 |
-
"
|
| 36 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 1.
|
| 41 |
-
"
|
| 42 |
-
"
|
| 43 |
-
"
|
| 44 |
-
"eval_samples_per_second": 43.856,
|
| 45 |
-
"eval_steps_per_second": 5.582,
|
| 46 |
-
"step": 42
|
| 47 |
-
},
|
| 48 |
-
{
|
| 49 |
-
"epoch": 1.1939393939393939,
|
| 50 |
-
"grad_norm": 86.90376281738281,
|
| 51 |
-
"learning_rate": 4.854368932038836e-06,
|
| 52 |
-
"loss": 0.7179,
|
| 53 |
"step": 50
|
| 54 |
},
|
| 55 |
{
|
| 56 |
-
"epoch": 1.
|
| 57 |
-
"grad_norm":
|
| 58 |
-
"learning_rate":
|
| 59 |
-
"loss": 0.
|
| 60 |
"step": 60
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"epoch":
|
| 64 |
-
"grad_norm":
|
| 65 |
-
"learning_rate":
|
| 66 |
-
"loss":
|
| 67 |
"step": 70
|
| 68 |
},
|
| 69 |
{
|
| 70 |
-
"epoch":
|
| 71 |
-
"
|
| 72 |
-
"
|
| 73 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
"step": 80
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch": 2.
|
| 78 |
-
"
|
| 79 |
-
"
|
| 80 |
-
"
|
| 81 |
-
"eval_samples_per_second": 56.169,
|
| 82 |
-
"eval_steps_per_second": 7.149,
|
| 83 |
-
"step": 84
|
| 84 |
-
},
|
| 85 |
-
{
|
| 86 |
-
"epoch": 2.1454545454545455,
|
| 87 |
-
"grad_norm": 232.57371520996094,
|
| 88 |
-
"learning_rate": 8.737864077669904e-06,
|
| 89 |
-
"loss": 1.323,
|
| 90 |
"step": 90
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"epoch": 2.
|
| 94 |
-
"grad_norm":
|
| 95 |
-
"learning_rate": 9.
|
| 96 |
-
"loss":
|
| 97 |
"step": 100
|
| 98 |
},
|
| 99 |
{
|
| 100 |
-
"epoch":
|
| 101 |
-
"
|
| 102 |
-
"
|
| 103 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 104 |
"step": 110
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch":
|
| 108 |
-
"grad_norm":
|
| 109 |
-
"learning_rate": 9.
|
| 110 |
-
"loss": 0.
|
| 111 |
"step": 120
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch": 3.
|
| 115 |
-
"
|
| 116 |
-
"
|
| 117 |
-
"
|
| 118 |
-
"eval_samples_per_second": 54.266,
|
| 119 |
-
"eval_steps_per_second": 6.907,
|
| 120 |
-
"step": 126
|
| 121 |
-
},
|
| 122 |
-
{
|
| 123 |
-
"epoch": 3.096969696969697,
|
| 124 |
-
"grad_norm": 40.3350715637207,
|
| 125 |
-
"learning_rate": 9.70715835140998e-06,
|
| 126 |
-
"loss": 0.6221,
|
| 127 |
"step": 130
|
| 128 |
},
|
| 129 |
{
|
| 130 |
-
"epoch":
|
| 131 |
-
"grad_norm":
|
| 132 |
-
"learning_rate": 9.
|
| 133 |
-
"loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 134 |
"step": 140
|
| 135 |
},
|
| 136 |
{
|
| 137 |
-
"epoch":
|
| 138 |
-
"grad_norm":
|
| 139 |
-
"learning_rate": 9.
|
| 140 |
-
"loss": 0.
|
| 141 |
"step": 150
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
-
"grad_norm":
|
| 146 |
-
"learning_rate": 9.
|
| 147 |
-
"loss": 0.
|
| 148 |
"step": 160
|
| 149 |
},
|
| 150 |
{
|
| 151 |
-
"epoch": 4.
|
| 152 |
-
"
|
| 153 |
-
"
|
| 154 |
-
"
|
| 155 |
-
"eval_samples_per_second": 55.655,
|
| 156 |
-
"eval_steps_per_second": 7.083,
|
| 157 |
-
"step": 168
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 4.048484848484849,
|
| 161 |
-
"grad_norm": 110.27919006347656,
|
| 162 |
-
"learning_rate": 9.273318872017354e-06,
|
| 163 |
-
"loss": 1.1381,
|
| 164 |
"step": 170
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch":
|
| 168 |
-
"
|
| 169 |
-
"
|
| 170 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 171 |
"step": 180
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch":
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 190
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch":
|
| 182 |
-
"grad_norm":
|
| 183 |
-
"learning_rate": 8.
|
| 184 |
-
"loss": 0.
|
| 185 |
"step": 200
|
| 186 |
},
|
| 187 |
{
|
| 188 |
-
"epoch":
|
| 189 |
-
"grad_norm":
|
| 190 |
-
"learning_rate": 8.
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 210
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch":
|
| 196 |
-
"eval_loss": 0.
|
| 197 |
-
"eval_macro_f1": 0.
|
| 198 |
-
"eval_runtime":
|
| 199 |
-
"eval_samples_per_second":
|
| 200 |
-
"eval_steps_per_second":
|
| 201 |
"step": 210
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch":
|
| 205 |
-
"grad_norm":
|
| 206 |
-
"learning_rate": 8.
|
| 207 |
-
"loss": 0.
|
| 208 |
"step": 220
|
| 209 |
},
|
| 210 |
{
|
| 211 |
-
"epoch":
|
| 212 |
-
"grad_norm":
|
| 213 |
-
"learning_rate": 8.
|
| 214 |
-
"loss": 0.
|
| 215 |
"step": 230
|
| 216 |
},
|
| 217 |
{
|
| 218 |
-
"epoch":
|
| 219 |
-
"grad_norm":
|
| 220 |
-
"learning_rate": 8.
|
| 221 |
-
"loss": 0.
|
| 222 |
"step": 240
|
| 223 |
},
|
| 224 |
{
|
| 225 |
-
"epoch":
|
| 226 |
-
"
|
| 227 |
-
"
|
| 228 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 229 |
"step": 250
|
| 230 |
},
|
| 231 |
{
|
| 232 |
-
"epoch":
|
| 233 |
-
"
|
| 234 |
-
"
|
| 235 |
-
"
|
| 236 |
-
"eval_samples_per_second": 53.223,
|
| 237 |
-
"eval_steps_per_second": 6.774,
|
| 238 |
-
"step": 252
|
| 239 |
-
},
|
| 240 |
-
{
|
| 241 |
-
"epoch": 6.193939393939394,
|
| 242 |
-
"grad_norm": 18.175823211669922,
|
| 243 |
-
"learning_rate": 8.29718004338395e-06,
|
| 244 |
-
"loss": 0.6752,
|
| 245 |
"step": 260
|
| 246 |
},
|
| 247 |
{
|
| 248 |
-
"epoch":
|
| 249 |
-
"grad_norm":
|
| 250 |
-
"learning_rate":
|
| 251 |
-
"loss": 0.
|
| 252 |
"step": 270
|
| 253 |
},
|
| 254 |
{
|
| 255 |
-
"epoch":
|
| 256 |
-
"grad_norm":
|
| 257 |
-
"learning_rate":
|
| 258 |
-
"loss":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 259 |
"step": 280
|
| 260 |
},
|
| 261 |
{
|
| 262 |
-
"epoch":
|
| 263 |
-
"grad_norm":
|
| 264 |
-
"learning_rate": 7.
|
| 265 |
-
"loss": 0.
|
| 266 |
"step": 290
|
| 267 |
},
|
| 268 |
{
|
| 269 |
-
"epoch":
|
| 270 |
-
"
|
| 271 |
-
"
|
| 272 |
-
"
|
| 273 |
-
"eval_samples_per_second": 53.91,
|
| 274 |
-
"eval_steps_per_second": 6.861,
|
| 275 |
-
"step": 294
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 7.1454545454545455,
|
| 279 |
-
"grad_norm": 41.31782531738281,
|
| 280 |
-
"learning_rate": 7.863340563991324e-06,
|
| 281 |
-
"loss": 0.3818,
|
| 282 |
"step": 300
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch":
|
| 286 |
-
"grad_norm":
|
| 287 |
-
"learning_rate": 7.
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 310
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch":
|
| 293 |
-
"
|
| 294 |
-
"
|
| 295 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 296 |
"step": 320
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch":
|
| 300 |
-
"grad_norm":
|
| 301 |
-
"learning_rate":
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 330
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch":
|
| 307 |
-
"
|
| 308 |
-
"
|
| 309 |
-
"
|
| 310 |
-
"eval_samples_per_second": 55.263,
|
| 311 |
-
"eval_steps_per_second": 7.033,
|
| 312 |
-
"step": 336
|
| 313 |
-
},
|
| 314 |
-
{
|
| 315 |
-
"epoch": 8.096969696969698,
|
| 316 |
-
"grad_norm": 9.571391105651855,
|
| 317 |
-
"learning_rate": 7.429501084598699e-06,
|
| 318 |
-
"loss": 0.3311,
|
| 319 |
"step": 340
|
| 320 |
},
|
| 321 |
-
{
|
| 322 |
-
"epoch": 8.33939393939394,
|
| 323 |
-
"grad_norm": 30.14655876159668,
|
| 324 |
-
"learning_rate": 7.321041214750543e-06,
|
| 325 |
-
"loss": 0.5367,
|
| 326 |
-
"step": 350
|
| 327 |
-
},
|
| 328 |
-
{
|
| 329 |
-
"epoch": 8.581818181818182,
|
| 330 |
-
"grad_norm": 125.38350677490234,
|
| 331 |
-
"learning_rate": 7.212581344902386e-06,
|
| 332 |
-
"loss": 0.4511,
|
| 333 |
-
"step": 360
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"epoch": 8.824242424242424,
|
| 337 |
-
"grad_norm": 283.20819091796875,
|
| 338 |
-
"learning_rate": 7.104121475054231e-06,
|
| 339 |
-
"loss": 0.5474,
|
| 340 |
-
"step": 370
|
| 341 |
-
},
|
| 342 |
-
{
|
| 343 |
-
"epoch": 9.0,
|
| 344 |
-
"eval_loss": 0.329227477312088,
|
| 345 |
-
"eval_macro_f1": 0.8817302125547928,
|
| 346 |
-
"eval_runtime": 5.984,
|
| 347 |
-
"eval_samples_per_second": 55.147,
|
| 348 |
-
"eval_steps_per_second": 7.019,
|
| 349 |
-
"step": 378
|
| 350 |
-
},
|
| 351 |
-
{
|
| 352 |
-
"epoch": 9.048484848484849,
|
| 353 |
-
"grad_norm": 31.927379608154297,
|
| 354 |
-
"learning_rate": 6.995661605206075e-06,
|
| 355 |
-
"loss": 0.3963,
|
| 356 |
-
"step": 380
|
| 357 |
-
},
|
| 358 |
-
{
|
| 359 |
-
"epoch": 9.290909090909091,
|
| 360 |
-
"grad_norm": 10.10098934173584,
|
| 361 |
-
"learning_rate": 6.887201735357918e-06,
|
| 362 |
-
"loss": 0.445,
|
| 363 |
-
"step": 390
|
| 364 |
-
},
|
| 365 |
-
{
|
| 366 |
-
"epoch": 9.533333333333333,
|
| 367 |
-
"grad_norm": 1.947770118713379,
|
| 368 |
-
"learning_rate": 6.778741865509761e-06,
|
| 369 |
-
"loss": 0.5947,
|
| 370 |
-
"step": 400
|
| 371 |
-
},
|
| 372 |
-
{
|
| 373 |
-
"epoch": 9.775757575757575,
|
| 374 |
-
"grad_norm": 54.11802673339844,
|
| 375 |
-
"learning_rate": 6.670281995661606e-06,
|
| 376 |
-
"loss": 0.6001,
|
| 377 |
-
"step": 410
|
| 378 |
-
},
|
| 379 |
{
|
| 380 |
"epoch": 10.0,
|
| 381 |
-
"grad_norm":
|
| 382 |
-
"learning_rate": 6.
|
| 383 |
-
"loss": 0.
|
| 384 |
-
"step":
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 10.0,
|
| 388 |
-
"eval_loss": 0.
|
| 389 |
-
"eval_macro_f1": 0.
|
| 390 |
-
"eval_runtime":
|
| 391 |
-
"eval_samples_per_second":
|
| 392 |
-
"eval_steps_per_second":
|
| 393 |
-
"step":
|
| 394 |
-
},
|
| 395 |
-
{
|
| 396 |
-
"epoch": 10.242424242424242,
|
| 397 |
-
"grad_norm": 0.22337216138839722,
|
| 398 |
-
"learning_rate": 6.453362255965293e-06,
|
| 399 |
-
"loss": 0.2798,
|
| 400 |
-
"step": 430
|
| 401 |
},
|
| 402 |
{
|
| 403 |
-
"epoch": 10.
|
| 404 |
-
"grad_norm":
|
| 405 |
-
"learning_rate": 6.
|
| 406 |
-
"loss": 0.
|
| 407 |
-
"step":
|
| 408 |
},
|
| 409 |
{
|
| 410 |
-
"epoch": 10.
|
| 411 |
-
"grad_norm":
|
| 412 |
-
"learning_rate": 6.
|
| 413 |
-
"loss": 0.
|
| 414 |
-
"step":
|
| 415 |
},
|
| 416 |
{
|
| 417 |
-
"epoch": 10.
|
| 418 |
-
"grad_norm":
|
| 419 |
-
"learning_rate": 6.
|
| 420 |
-
"loss": 0.
|
| 421 |
-
"step":
|
| 422 |
},
|
| 423 |
{
|
| 424 |
"epoch": 11.0,
|
| 425 |
-
"eval_loss": 0.
|
| 426 |
-
"eval_macro_f1": 0.
|
| 427 |
-
"eval_runtime":
|
| 428 |
-
"eval_samples_per_second":
|
| 429 |
-
"eval_steps_per_second":
|
| 430 |
-
"step":
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 11.193939393939393,
|
| 434 |
-
"grad_norm": 63.143218994140625,
|
| 435 |
-
"learning_rate": 6.019522776572668e-06,
|
| 436 |
-
"loss": 0.6075,
|
| 437 |
-
"step": 470
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 11.436363636363636,
|
| 441 |
-
"grad_norm": 7.950187683105469,
|
| 442 |
-
"learning_rate": 5.911062906724513e-06,
|
| 443 |
-
"loss": 0.2654,
|
| 444 |
-
"step": 480
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 11.67878787878788,
|
| 448 |
-
"grad_norm": 82.30758666992188,
|
| 449 |
-
"learning_rate": 5.802603036876356e-06,
|
| 450 |
-
"loss": 0.2474,
|
| 451 |
-
"step": 490
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 11.921212121212122,
|
| 455 |
-
"grad_norm": 7.340689182281494,
|
| 456 |
-
"learning_rate": 5.6941431670282e-06,
|
| 457 |
-
"loss": 0.2405,
|
| 458 |
-
"step": 500
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 12.0,
|
| 462 |
-
"eval_loss": 0.3316686451435089,
|
| 463 |
-
"eval_macro_f1": 0.9209216589861751,
|
| 464 |
-
"eval_runtime": 5.8916,
|
| 465 |
-
"eval_samples_per_second": 56.012,
|
| 466 |
-
"eval_steps_per_second": 7.129,
|
| 467 |
-
"step": 504
|
| 468 |
-
},
|
| 469 |
-
{
|
| 470 |
-
"epoch": 12.145454545454545,
|
| 471 |
-
"grad_norm": 37.141632080078125,
|
| 472 |
-
"learning_rate": 5.585683297180043e-06,
|
| 473 |
-
"loss": 0.3349,
|
| 474 |
-
"step": 510
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 12.387878787878789,
|
| 478 |
-
"grad_norm": 34.024383544921875,
|
| 479 |
-
"learning_rate": 5.477223427331888e-06,
|
| 480 |
-
"loss": 0.1742,
|
| 481 |
-
"step": 520
|
| 482 |
-
},
|
| 483 |
-
{
|
| 484 |
-
"epoch": 12.63030303030303,
|
| 485 |
-
"grad_norm": 46.10781478881836,
|
| 486 |
-
"learning_rate": 5.368763557483731e-06,
|
| 487 |
-
"loss": 0.2115,
|
| 488 |
-
"step": 530
|
| 489 |
-
},
|
| 490 |
-
{
|
| 491 |
-
"epoch": 12.872727272727273,
|
| 492 |
-
"grad_norm": 126.67015838623047,
|
| 493 |
-
"learning_rate": 5.260303687635575e-06,
|
| 494 |
-
"loss": 0.8788,
|
| 495 |
-
"step": 540
|
| 496 |
-
},
|
| 497 |
-
{
|
| 498 |
-
"epoch": 13.0,
|
| 499 |
-
"eval_loss": 0.37741926312446594,
|
| 500 |
-
"eval_macro_f1": 0.9058106453305834,
|
| 501 |
-
"eval_runtime": 6.6329,
|
| 502 |
-
"eval_samples_per_second": 49.752,
|
| 503 |
-
"eval_steps_per_second": 6.332,
|
| 504 |
-
"step": 546
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 13.096969696969698,
|
| 508 |
-
"grad_norm": 35.44662094116211,
|
| 509 |
-
"learning_rate": 5.151843817787418e-06,
|
| 510 |
-
"loss": 0.5591,
|
| 511 |
-
"step": 550
|
| 512 |
-
},
|
| 513 |
-
{
|
| 514 |
-
"epoch": 13.33939393939394,
|
| 515 |
-
"grad_norm": 57.34544372558594,
|
| 516 |
-
"learning_rate": 5.043383947939263e-06,
|
| 517 |
-
"loss": 0.213,
|
| 518 |
-
"step": 560
|
| 519 |
-
},
|
| 520 |
-
{
|
| 521 |
-
"epoch": 13.581818181818182,
|
| 522 |
-
"grad_norm": 15.285223960876465,
|
| 523 |
-
"learning_rate": 4.934924078091107e-06,
|
| 524 |
-
"loss": 0.203,
|
| 525 |
-
"step": 570
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 13.824242424242424,
|
| 529 |
-
"grad_norm": 28.99003028869629,
|
| 530 |
-
"learning_rate": 4.82646420824295e-06,
|
| 531 |
-
"loss": 0.4163,
|
| 532 |
-
"step": 580
|
| 533 |
-
},
|
| 534 |
-
{
|
| 535 |
-
"epoch": 14.0,
|
| 536 |
-
"eval_loss": 0.39865490794181824,
|
| 537 |
-
"eval_macro_f1": 0.9196508840275697,
|
| 538 |
-
"eval_runtime": 5.8701,
|
| 539 |
-
"eval_samples_per_second": 56.217,
|
| 540 |
-
"eval_steps_per_second": 7.155,
|
| 541 |
-
"step": 588
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 14.048484848484849,
|
| 545 |
-
"grad_norm": 346.8255310058594,
|
| 546 |
-
"learning_rate": 4.718004338394794e-06,
|
| 547 |
-
"loss": 0.26,
|
| 548 |
-
"step": 590
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 14.290909090909091,
|
| 552 |
-
"grad_norm": 38.04654312133789,
|
| 553 |
-
"learning_rate": 4.609544468546638e-06,
|
| 554 |
-
"loss": 0.3813,
|
| 555 |
-
"step": 600
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 14.533333333333333,
|
| 559 |
-
"grad_norm": 34.71643829345703,
|
| 560 |
-
"learning_rate": 4.501084598698482e-06,
|
| 561 |
-
"loss": 0.0974,
|
| 562 |
-
"step": 610
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 14.775757575757575,
|
| 566 |
-
"grad_norm": 34.031890869140625,
|
| 567 |
-
"learning_rate": 4.392624728850326e-06,
|
| 568 |
-
"loss": 0.4881,
|
| 569 |
-
"step": 620
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 15.0,
|
| 573 |
-
"grad_norm": 0.0002771662548184395,
|
| 574 |
-
"learning_rate": 4.284164859002169e-06,
|
| 575 |
-
"loss": 0.4126,
|
| 576 |
-
"step": 630
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 15.0,
|
| 580 |
-
"eval_loss": 0.35451531410217285,
|
| 581 |
-
"eval_macro_f1": 0.9235679411519468,
|
| 582 |
-
"eval_runtime": 6.0428,
|
| 583 |
-
"eval_samples_per_second": 54.611,
|
| 584 |
-
"eval_steps_per_second": 6.95,
|
| 585 |
-
"step": 630
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 15.242424242424242,
|
| 589 |
-
"grad_norm": 82.48748779296875,
|
| 590 |
-
"learning_rate": 4.175704989154013e-06,
|
| 591 |
-
"loss": 0.4444,
|
| 592 |
-
"step": 640
|
| 593 |
-
},
|
| 594 |
-
{
|
| 595 |
-
"epoch": 15.484848484848484,
|
| 596 |
-
"grad_norm": 0.2618753910064697,
|
| 597 |
-
"learning_rate": 4.067245119305857e-06,
|
| 598 |
-
"loss": 0.2083,
|
| 599 |
-
"step": 650
|
| 600 |
-
},
|
| 601 |
-
{
|
| 602 |
-
"epoch": 15.727272727272727,
|
| 603 |
-
"grad_norm": 98.34405517578125,
|
| 604 |
-
"learning_rate": 3.958785249457701e-06,
|
| 605 |
-
"loss": 0.4785,
|
| 606 |
-
"step": 660
|
| 607 |
-
},
|
| 608 |
-
{
|
| 609 |
-
"epoch": 15.969696969696969,
|
| 610 |
-
"grad_norm": 0.37142229080200195,
|
| 611 |
-
"learning_rate": 3.8503253796095445e-06,
|
| 612 |
-
"loss": 0.1583,
|
| 613 |
-
"step": 670
|
| 614 |
-
},
|
| 615 |
-
{
|
| 616 |
-
"epoch": 16.0,
|
| 617 |
-
"eval_loss": 0.38117873668670654,
|
| 618 |
-
"eval_macro_f1": 0.9268860086407444,
|
| 619 |
-
"eval_runtime": 6.9311,
|
| 620 |
-
"eval_samples_per_second": 47.612,
|
| 621 |
-
"eval_steps_per_second": 6.06,
|
| 622 |
-
"step": 672
|
| 623 |
-
},
|
| 624 |
-
{
|
| 625 |
-
"epoch": 16.193939393939395,
|
| 626 |
-
"grad_norm": 40.54330825805664,
|
| 627 |
-
"learning_rate": 3.741865509761389e-06,
|
| 628 |
-
"loss": 0.0774,
|
| 629 |
-
"step": 680
|
| 630 |
-
},
|
| 631 |
-
{
|
| 632 |
-
"epoch": 16.436363636363637,
|
| 633 |
-
"grad_norm": 0.23675695061683655,
|
| 634 |
-
"learning_rate": 3.6334056399132324e-06,
|
| 635 |
-
"loss": 0.1639,
|
| 636 |
-
"step": 690
|
| 637 |
-
},
|
| 638 |
-
{
|
| 639 |
-
"epoch": 16.67878787878788,
|
| 640 |
-
"grad_norm": 47.12529373168945,
|
| 641 |
-
"learning_rate": 3.5249457700650764e-06,
|
| 642 |
-
"loss": 0.306,
|
| 643 |
-
"step": 700
|
| 644 |
-
},
|
| 645 |
-
{
|
| 646 |
-
"epoch": 16.921212121212122,
|
| 647 |
-
"grad_norm": 0.3993530571460724,
|
| 648 |
-
"learning_rate": 3.41648590021692e-06,
|
| 649 |
-
"loss": 0.2376,
|
| 650 |
-
"step": 710
|
| 651 |
-
},
|
| 652 |
-
{
|
| 653 |
-
"epoch": 17.0,
|
| 654 |
-
"eval_loss": 0.4087267816066742,
|
| 655 |
-
"eval_macro_f1": 0.9295990205081115,
|
| 656 |
-
"eval_runtime": 6.1306,
|
| 657 |
-
"eval_samples_per_second": 53.828,
|
| 658 |
-
"eval_steps_per_second": 6.851,
|
| 659 |
-
"step": 714
|
| 660 |
-
},
|
| 661 |
-
{
|
| 662 |
-
"epoch": 17.145454545454545,
|
| 663 |
-
"grad_norm": 0.34205177426338196,
|
| 664 |
-
"learning_rate": 3.308026030368764e-06,
|
| 665 |
-
"loss": 0.0332,
|
| 666 |
-
"step": 720
|
| 667 |
-
},
|
| 668 |
-
{
|
| 669 |
-
"epoch": 17.387878787878787,
|
| 670 |
-
"grad_norm": 0.5112647414207458,
|
| 671 |
-
"learning_rate": 3.1995661605206075e-06,
|
| 672 |
-
"loss": 0.1332,
|
| 673 |
-
"step": 730
|
| 674 |
-
},
|
| 675 |
-
{
|
| 676 |
-
"epoch": 17.63030303030303,
|
| 677 |
-
"grad_norm": 120.2950439453125,
|
| 678 |
-
"learning_rate": 3.0911062906724515e-06,
|
| 679 |
-
"loss": 0.2503,
|
| 680 |
-
"step": 740
|
| 681 |
-
},
|
| 682 |
-
{
|
| 683 |
-
"epoch": 17.87272727272727,
|
| 684 |
-
"grad_norm": 223.04759216308594,
|
| 685 |
-
"learning_rate": 2.982646420824295e-06,
|
| 686 |
-
"loss": 0.2703,
|
| 687 |
-
"step": 750
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"epoch": 18.0,
|
| 691 |
-
"eval_loss": 0.43362897634506226,
|
| 692 |
-
"eval_macro_f1": 0.9264924264924266,
|
| 693 |
-
"eval_runtime": 5.861,
|
| 694 |
-
"eval_samples_per_second": 56.305,
|
| 695 |
-
"eval_steps_per_second": 7.166,
|
| 696 |
-
"step": 756
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"epoch": 18.096969696969698,
|
| 700 |
-
"grad_norm": 54.66193771362305,
|
| 701 |
-
"learning_rate": 2.874186550976139e-06,
|
| 702 |
-
"loss": 0.1274,
|
| 703 |
-
"step": 760
|
| 704 |
-
},
|
| 705 |
-
{
|
| 706 |
-
"epoch": 18.33939393939394,
|
| 707 |
-
"grad_norm": 54.846466064453125,
|
| 708 |
-
"learning_rate": 2.765726681127983e-06,
|
| 709 |
-
"loss": 0.2751,
|
| 710 |
-
"step": 770
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 18.581818181818182,
|
| 714 |
-
"grad_norm": 53.97863006591797,
|
| 715 |
-
"learning_rate": 2.6572668112798266e-06,
|
| 716 |
-
"loss": 0.359,
|
| 717 |
-
"step": 780
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 18.824242424242424,
|
| 721 |
-
"grad_norm": 81.63549041748047,
|
| 722 |
-
"learning_rate": 2.5488069414316706e-06,
|
| 723 |
-
"loss": 0.1819,
|
| 724 |
-
"step": 790
|
| 725 |
-
},
|
| 726 |
-
{
|
| 727 |
-
"epoch": 19.0,
|
| 728 |
-
"eval_loss": 0.3480012118816376,
|
| 729 |
-
"eval_macro_f1": 0.9236528192931639,
|
| 730 |
-
"eval_runtime": 7.0471,
|
| 731 |
-
"eval_samples_per_second": 46.828,
|
| 732 |
-
"eval_steps_per_second": 5.96,
|
| 733 |
-
"step": 798
|
| 734 |
-
},
|
| 735 |
-
{
|
| 736 |
-
"epoch": 19.048484848484847,
|
| 737 |
-
"grad_norm": 41.54087448120117,
|
| 738 |
-
"learning_rate": 2.440347071583514e-06,
|
| 739 |
-
"loss": 0.6373,
|
| 740 |
-
"step": 800
|
| 741 |
-
},
|
| 742 |
-
{
|
| 743 |
-
"epoch": 19.29090909090909,
|
| 744 |
-
"grad_norm": 9.001028060913086,
|
| 745 |
-
"learning_rate": 2.331887201735358e-06,
|
| 746 |
-
"loss": 0.2971,
|
| 747 |
-
"step": 810
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 19.533333333333335,
|
| 751 |
-
"grad_norm": 114.6279525756836,
|
| 752 |
-
"learning_rate": 2.2234273318872017e-06,
|
| 753 |
-
"loss": 0.1943,
|
| 754 |
-
"step": 820
|
| 755 |
-
},
|
| 756 |
-
{
|
| 757 |
-
"epoch": 19.775757575757577,
|
| 758 |
-
"grad_norm": 18.022676467895508,
|
| 759 |
-
"learning_rate": 2.1149674620390457e-06,
|
| 760 |
-
"loss": 0.1207,
|
| 761 |
-
"step": 830
|
| 762 |
-
},
|
| 763 |
-
{
|
| 764 |
-
"epoch": 20.0,
|
| 765 |
-
"grad_norm": 0.0001882202341221273,
|
| 766 |
-
"learning_rate": 2.0065075921908892e-06,
|
| 767 |
-
"loss": 0.1324,
|
| 768 |
-
"step": 840
|
| 769 |
-
},
|
| 770 |
-
{
|
| 771 |
-
"epoch": 20.0,
|
| 772 |
-
"eval_loss": 0.4493299424648285,
|
| 773 |
-
"eval_macro_f1": 0.9384902143522833,
|
| 774 |
-
"eval_runtime": 6.0147,
|
| 775 |
-
"eval_samples_per_second": 54.865,
|
| 776 |
-
"eval_steps_per_second": 6.983,
|
| 777 |
-
"step": 840
|
| 778 |
-
},
|
| 779 |
-
{
|
| 780 |
-
"epoch": 20.242424242424242,
|
| 781 |
-
"grad_norm": 13.740226745605469,
|
| 782 |
-
"learning_rate": 1.8980477223427332e-06,
|
| 783 |
-
"loss": 0.294,
|
| 784 |
-
"step": 850
|
| 785 |
-
},
|
| 786 |
-
{
|
| 787 |
-
"epoch": 20.484848484848484,
|
| 788 |
-
"grad_norm": 6.870513916015625,
|
| 789 |
-
"learning_rate": 1.7895878524945772e-06,
|
| 790 |
-
"loss": 0.1323,
|
| 791 |
-
"step": 860
|
| 792 |
-
},
|
| 793 |
-
{
|
| 794 |
-
"epoch": 20.727272727272727,
|
| 795 |
-
"grad_norm": 2.74729585647583,
|
| 796 |
-
"learning_rate": 1.681127982646421e-06,
|
| 797 |
-
"loss": 0.019,
|
| 798 |
-
"step": 870
|
| 799 |
-
},
|
| 800 |
-
{
|
| 801 |
-
"epoch": 20.96969696969697,
|
| 802 |
-
"grad_norm": 17.18338966369629,
|
| 803 |
-
"learning_rate": 1.572668112798265e-06,
|
| 804 |
-
"loss": 0.1312,
|
| 805 |
-
"step": 880
|
| 806 |
-
},
|
| 807 |
-
{
|
| 808 |
-
"epoch": 21.0,
|
| 809 |
-
"eval_loss": 0.40448498725891113,
|
| 810 |
-
"eval_macro_f1": 0.9384902143522833,
|
| 811 |
-
"eval_runtime": 6.133,
|
| 812 |
-
"eval_samples_per_second": 53.808,
|
| 813 |
-
"eval_steps_per_second": 6.848,
|
| 814 |
-
"step": 882
|
| 815 |
-
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 21.193939393939395,
|
| 818 |
-
"grad_norm": 108.9833755493164,
|
| 819 |
-
"learning_rate": 1.4642082429501087e-06,
|
| 820 |
-
"loss": 0.2499,
|
| 821 |
-
"step": 890
|
| 822 |
-
},
|
| 823 |
-
{
|
| 824 |
-
"epoch": 21.436363636363637,
|
| 825 |
-
"grad_norm": 0.291847825050354,
|
| 826 |
-
"learning_rate": 1.3557483731019525e-06,
|
| 827 |
-
"loss": 0.1708,
|
| 828 |
-
"step": 900
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 21.67878787878788,
|
| 832 |
-
"grad_norm": 35.08168029785156,
|
| 833 |
-
"learning_rate": 1.2472885032537963e-06,
|
| 834 |
-
"loss": 0.0802,
|
| 835 |
-
"step": 910
|
| 836 |
-
},
|
| 837 |
-
{
|
| 838 |
-
"epoch": 21.921212121212122,
|
| 839 |
-
"grad_norm": 0.05401836335659027,
|
| 840 |
-
"learning_rate": 1.13882863340564e-06,
|
| 841 |
-
"loss": 0.1662,
|
| 842 |
-
"step": 920
|
| 843 |
-
},
|
| 844 |
-
{
|
| 845 |
-
"epoch": 22.0,
|
| 846 |
-
"eval_loss": 0.3166828453540802,
|
| 847 |
-
"eval_macro_f1": 0.9417692129092176,
|
| 848 |
-
"eval_runtime": 6.0442,
|
| 849 |
-
"eval_samples_per_second": 54.598,
|
| 850 |
-
"eval_steps_per_second": 6.949,
|
| 851 |
-
"step": 924
|
| 852 |
}
|
| 853 |
],
|
| 854 |
"logging_steps": 10,
|
| 855 |
-
"max_steps":
|
| 856 |
"num_input_tokens_seen": 0,
|
| 857 |
"num_train_epochs": 25,
|
| 858 |
"save_steps": 500,
|
|
@@ -868,7 +392,7 @@
|
|
| 868 |
"attributes": {}
|
| 869 |
}
|
| 870 |
},
|
| 871 |
-
"total_flos":
|
| 872 |
"train_batch_size": 8,
|
| 873 |
"trial_name": null,
|
| 874 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 385,
|
| 3 |
+
"best_metric": 0.9489664082687339,
|
| 4 |
+
"best_model_checkpoint": "Crosswalk/dinov2/checkpoint-385",
|
| 5 |
+
"epoch": 11.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 385,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.28776978417266186,
|
| 14 |
+
"grad_norm": 25.290170669555664,
|
| 15 |
+
"learning_rate": 1.0227272727272729e-06,
|
| 16 |
+
"loss": 0.7613,
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.5755395683453237,
|
| 21 |
+
"grad_norm": 19.748838424682617,
|
| 22 |
+
"learning_rate": 2.1590909090909092e-06,
|
| 23 |
+
"loss": 0.4653,
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.8633093525179856,
|
| 28 |
+
"grad_norm": 642.5625,
|
| 29 |
+
"learning_rate": 3.2954545454545456e-06,
|
| 30 |
+
"loss": 0.3536,
|
| 31 |
"step": 30
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 1.0,
|
| 35 |
+
"eval_loss": 0.3255312740802765,
|
| 36 |
+
"eval_macro_f1": 0.8774975492380359,
|
| 37 |
+
"eval_runtime": 26.9331,
|
| 38 |
+
"eval_samples_per_second": 8.8,
|
| 39 |
+
"eval_steps_per_second": 1.114,
|
| 40 |
+
"step": 35
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 1.143884892086331,
|
| 44 |
+
"grad_norm": 18.897483825683594,
|
| 45 |
+
"learning_rate": 4.4318181818181824e-06,
|
| 46 |
+
"loss": 0.3086,
|
| 47 |
"step": 40
|
| 48 |
},
|
| 49 |
{
|
| 50 |
+
"epoch": 1.4316546762589928,
|
| 51 |
+
"grad_norm": 7.919783115386963,
|
| 52 |
+
"learning_rate": 5.568181818181818e-06,
|
| 53 |
+
"loss": 0.1994,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 54 |
"step": 50
|
| 55 |
},
|
| 56 |
{
|
| 57 |
+
"epoch": 1.7194244604316546,
|
| 58 |
+
"grad_norm": 16.423315048217773,
|
| 59 |
+
"learning_rate": 6.704545454545454e-06,
|
| 60 |
+
"loss": 0.2462,
|
| 61 |
"step": 60
|
| 62 |
},
|
| 63 |
{
|
| 64 |
+
"epoch": 2.0,
|
| 65 |
+
"grad_norm": 71.16481018066406,
|
| 66 |
+
"learning_rate": 7.840909090909091e-06,
|
| 67 |
+
"loss": 0.2856,
|
| 68 |
"step": 70
|
| 69 |
},
|
| 70 |
{
|
| 71 |
+
"epoch": 2.0,
|
| 72 |
+
"eval_loss": 0.30485469102859497,
|
| 73 |
+
"eval_macro_f1": 0.8690491043579004,
|
| 74 |
+
"eval_runtime": 29.715,
|
| 75 |
+
"eval_samples_per_second": 7.976,
|
| 76 |
+
"eval_steps_per_second": 1.01,
|
| 77 |
+
"step": 70
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 2.287769784172662,
|
| 81 |
+
"grad_norm": 30.668006896972656,
|
| 82 |
+
"learning_rate": 8.977272727272727e-06,
|
| 83 |
+
"loss": 0.1576,
|
| 84 |
"step": 80
|
| 85 |
},
|
| 86 |
{
|
| 87 |
+
"epoch": 2.5755395683453237,
|
| 88 |
+
"grad_norm": 7.689505100250244,
|
| 89 |
+
"learning_rate": 9.987293519695045e-06,
|
| 90 |
+
"loss": 0.212,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 91 |
"step": 90
|
| 92 |
},
|
| 93 |
{
|
| 94 |
+
"epoch": 2.8633093525179856,
|
| 95 |
+
"grad_norm": 30.23891830444336,
|
| 96 |
+
"learning_rate": 9.86022871664549e-06,
|
| 97 |
+
"loss": 0.3253,
|
| 98 |
"step": 100
|
| 99 |
},
|
| 100 |
{
|
| 101 |
+
"epoch": 3.0,
|
| 102 |
+
"eval_loss": 0.2832421362400055,
|
| 103 |
+
"eval_macro_f1": 0.892023399486086,
|
| 104 |
+
"eval_runtime": 29.6551,
|
| 105 |
+
"eval_samples_per_second": 7.992,
|
| 106 |
+
"eval_steps_per_second": 1.012,
|
| 107 |
+
"step": 105
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 3.143884892086331,
|
| 111 |
+
"grad_norm": 18.362842559814453,
|
| 112 |
+
"learning_rate": 9.733163913595934e-06,
|
| 113 |
+
"loss": 0.2215,
|
| 114 |
"step": 110
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 3.431654676258993,
|
| 118 |
+
"grad_norm": 44.57633590698242,
|
| 119 |
+
"learning_rate": 9.60609911054638e-06,
|
| 120 |
+
"loss": 0.1955,
|
| 121 |
"step": 120
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 3.7194244604316546,
|
| 125 |
+
"grad_norm": 10.040678024291992,
|
| 126 |
+
"learning_rate": 9.479034307496824e-06,
|
| 127 |
+
"loss": 0.2095,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 128 |
"step": 130
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 4.0,
|
| 132 |
+
"grad_norm": 14.543230056762695,
|
| 133 |
+
"learning_rate": 9.35196950444727e-06,
|
| 134 |
+
"loss": 0.2134,
|
| 135 |
+
"step": 140
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 4.0,
|
| 139 |
+
"eval_loss": 0.28937143087387085,
|
| 140 |
+
"eval_macro_f1": 0.9314632735685368,
|
| 141 |
+
"eval_runtime": 30.5498,
|
| 142 |
+
"eval_samples_per_second": 7.758,
|
| 143 |
+
"eval_steps_per_second": 0.982,
|
| 144 |
"step": 140
|
| 145 |
},
|
| 146 |
{
|
| 147 |
+
"epoch": 4.287769784172662,
|
| 148 |
+
"grad_norm": 41.635799407958984,
|
| 149 |
+
"learning_rate": 9.224904701397714e-06,
|
| 150 |
+
"loss": 0.0965,
|
| 151 |
"step": 150
|
| 152 |
},
|
| 153 |
{
|
| 154 |
+
"epoch": 4.575539568345324,
|
| 155 |
+
"grad_norm": 21.393756866455078,
|
| 156 |
+
"learning_rate": 9.09783989834816e-06,
|
| 157 |
+
"loss": 0.1777,
|
| 158 |
"step": 160
|
| 159 |
},
|
| 160 |
{
|
| 161 |
+
"epoch": 4.863309352517986,
|
| 162 |
+
"grad_norm": 18.873945236206055,
|
| 163 |
+
"learning_rate": 8.970775095298603e-06,
|
| 164 |
+
"loss": 0.2155,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 165 |
"step": 170
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 5.0,
|
| 169 |
+
"eval_loss": 0.26767683029174805,
|
| 170 |
+
"eval_macro_f1": 0.9277555631264681,
|
| 171 |
+
"eval_runtime": 26.6297,
|
| 172 |
+
"eval_samples_per_second": 8.9,
|
| 173 |
+
"eval_steps_per_second": 1.127,
|
| 174 |
+
"step": 175
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 5.143884892086331,
|
| 178 |
+
"grad_norm": 6.233393669128418,
|
| 179 |
+
"learning_rate": 8.843710292249047e-06,
|
| 180 |
+
"loss": 0.1474,
|
| 181 |
"step": 180
|
| 182 |
},
|
| 183 |
{
|
| 184 |
+
"epoch": 5.431654676258993,
|
| 185 |
+
"grad_norm": 20.907960891723633,
|
| 186 |
+
"learning_rate": 8.716645489199493e-06,
|
| 187 |
+
"loss": 0.1279,
|
| 188 |
"step": 190
|
| 189 |
},
|
| 190 |
{
|
| 191 |
+
"epoch": 5.719424460431655,
|
| 192 |
+
"grad_norm": 17.63884162902832,
|
| 193 |
+
"learning_rate": 8.589580686149937e-06,
|
| 194 |
+
"loss": 0.211,
|
| 195 |
"step": 200
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 6.0,
|
| 199 |
+
"grad_norm": 20.169618606567383,
|
| 200 |
+
"learning_rate": 8.462515883100381e-06,
|
| 201 |
+
"loss": 0.108,
|
| 202 |
"step": 210
|
| 203 |
},
|
| 204 |
{
|
| 205 |
+
"epoch": 6.0,
|
| 206 |
+
"eval_loss": 0.2715422213077545,
|
| 207 |
+
"eval_macro_f1": 0.9110821288835689,
|
| 208 |
+
"eval_runtime": 32.5783,
|
| 209 |
+
"eval_samples_per_second": 7.275,
|
| 210 |
+
"eval_steps_per_second": 0.921,
|
| 211 |
"step": 210
|
| 212 |
},
|
| 213 |
{
|
| 214 |
+
"epoch": 6.287769784172662,
|
| 215 |
+
"grad_norm": 6.99655294418335,
|
| 216 |
+
"learning_rate": 8.335451080050827e-06,
|
| 217 |
+
"loss": 0.1142,
|
| 218 |
"step": 220
|
| 219 |
},
|
| 220 |
{
|
| 221 |
+
"epoch": 6.575539568345324,
|
| 222 |
+
"grad_norm": 20.068387985229492,
|
| 223 |
+
"learning_rate": 8.20838627700127e-06,
|
| 224 |
+
"loss": 0.2136,
|
| 225 |
"step": 230
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 6.863309352517986,
|
| 229 |
+
"grad_norm": 12.439704895019531,
|
| 230 |
+
"learning_rate": 8.081321473951716e-06,
|
| 231 |
+
"loss": 0.1461,
|
| 232 |
"step": 240
|
| 233 |
},
|
| 234 |
{
|
| 235 |
+
"epoch": 7.0,
|
| 236 |
+
"eval_loss": 0.49200543761253357,
|
| 237 |
+
"eval_macro_f1": 0.8094617047505047,
|
| 238 |
+
"eval_runtime": 38.0217,
|
| 239 |
+
"eval_samples_per_second": 6.233,
|
| 240 |
+
"eval_steps_per_second": 0.789,
|
| 241 |
+
"step": 245
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 7.143884892086331,
|
| 245 |
+
"grad_norm": 4.405595302581787,
|
| 246 |
+
"learning_rate": 7.95425667090216e-06,
|
| 247 |
+
"loss": 0.1535,
|
| 248 |
"step": 250
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 7.431654676258993,
|
| 252 |
+
"grad_norm": 7.879202365875244,
|
| 253 |
+
"learning_rate": 7.827191867852606e-06,
|
| 254 |
+
"loss": 0.2075,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 255 |
"step": 260
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 7.719424460431655,
|
| 259 |
+
"grad_norm": 4.8070268630981445,
|
| 260 |
+
"learning_rate": 7.70012706480305e-06,
|
| 261 |
+
"loss": 0.1155,
|
| 262 |
"step": 270
|
| 263 |
},
|
| 264 |
{
|
| 265 |
+
"epoch": 8.0,
|
| 266 |
+
"grad_norm": 12.835691452026367,
|
| 267 |
+
"learning_rate": 7.573062261753494e-06,
|
| 268 |
+
"loss": 0.1726,
|
| 269 |
+
"step": 280
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 8.0,
|
| 273 |
+
"eval_loss": 0.2971098721027374,
|
| 274 |
+
"eval_macro_f1": 0.9273962481754455,
|
| 275 |
+
"eval_runtime": 30.7621,
|
| 276 |
+
"eval_samples_per_second": 7.704,
|
| 277 |
+
"eval_steps_per_second": 0.975,
|
| 278 |
"step": 280
|
| 279 |
},
|
| 280 |
{
|
| 281 |
+
"epoch": 8.287769784172662,
|
| 282 |
+
"grad_norm": 10.0529146194458,
|
| 283 |
+
"learning_rate": 7.44599745870394e-06,
|
| 284 |
+
"loss": 0.0781,
|
| 285 |
"step": 290
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 8.575539568345324,
|
| 289 |
+
"grad_norm": 8.834739685058594,
|
| 290 |
+
"learning_rate": 7.318932655654384e-06,
|
| 291 |
+
"loss": 0.2262,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 292 |
"step": 300
|
| 293 |
},
|
| 294 |
{
|
| 295 |
+
"epoch": 8.863309352517986,
|
| 296 |
+
"grad_norm": 8.839664459228516,
|
| 297 |
+
"learning_rate": 7.191867852604829e-06,
|
| 298 |
+
"loss": 0.082,
|
| 299 |
"step": 310
|
| 300 |
},
|
| 301 |
{
|
| 302 |
+
"epoch": 9.0,
|
| 303 |
+
"eval_loss": 0.38130030035972595,
|
| 304 |
+
"eval_macro_f1": 0.8962043795620438,
|
| 305 |
+
"eval_runtime": 32.1353,
|
| 306 |
+
"eval_samples_per_second": 7.375,
|
| 307 |
+
"eval_steps_per_second": 0.934,
|
| 308 |
+
"step": 315
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 9.14388489208633,
|
| 312 |
+
"grad_norm": 14.185478210449219,
|
| 313 |
+
"learning_rate": 7.064803049555273e-06,
|
| 314 |
+
"loss": 0.1111,
|
| 315 |
"step": 320
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 9.431654676258994,
|
| 319 |
+
"grad_norm": 2.165837049484253,
|
| 320 |
+
"learning_rate": 6.937738246505718e-06,
|
| 321 |
+
"loss": 0.0988,
|
| 322 |
"step": 330
|
| 323 |
},
|
| 324 |
{
|
| 325 |
+
"epoch": 9.719424460431654,
|
| 326 |
+
"grad_norm": 1.7402962446212769,
|
| 327 |
+
"learning_rate": 6.810673443456163e-06,
|
| 328 |
+
"loss": 0.0941,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 329 |
"step": 340
|
| 330 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 331 |
{
|
| 332 |
"epoch": 10.0,
|
| 333 |
+
"grad_norm": 35.40852737426758,
|
| 334 |
+
"learning_rate": 6.683608640406608e-06,
|
| 335 |
+
"loss": 0.0827,
|
| 336 |
+
"step": 350
|
| 337 |
},
|
| 338 |
{
|
| 339 |
"epoch": 10.0,
|
| 340 |
+
"eval_loss": 0.34335753321647644,
|
| 341 |
+
"eval_macro_f1": 0.9318475916606758,
|
| 342 |
+
"eval_runtime": 34.461,
|
| 343 |
+
"eval_samples_per_second": 6.877,
|
| 344 |
+
"eval_steps_per_second": 0.871,
|
| 345 |
+
"step": 350
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 10.287769784172662,
|
| 349 |
+
"grad_norm": 21.371692657470703,
|
| 350 |
+
"learning_rate": 6.556543837357052e-06,
|
| 351 |
+
"loss": 0.0953,
|
| 352 |
+
"step": 360
|
| 353 |
},
|
| 354 |
{
|
| 355 |
+
"epoch": 10.575539568345324,
|
| 356 |
+
"grad_norm": 0.9700206518173218,
|
| 357 |
+
"learning_rate": 6.4294790343074975e-06,
|
| 358 |
+
"loss": 0.0855,
|
| 359 |
+
"step": 370
|
| 360 |
},
|
| 361 |
{
|
| 362 |
+
"epoch": 10.863309352517986,
|
| 363 |
+
"grad_norm": 11.236717224121094,
|
| 364 |
+
"learning_rate": 6.3024142312579415e-06,
|
| 365 |
+
"loss": 0.0853,
|
| 366 |
+
"step": 380
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 11.0,
|
| 370 |
+
"eval_loss": 0.27990925312042236,
|
| 371 |
+
"eval_macro_f1": 0.9489664082687339,
|
| 372 |
+
"eval_runtime": 32.6676,
|
| 373 |
+
"eval_samples_per_second": 7.255,
|
| 374 |
+
"eval_steps_per_second": 0.918,
|
| 375 |
+
"step": 385
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 376 |
}
|
| 377 |
],
|
| 378 |
"logging_steps": 10,
|
| 379 |
+
"max_steps": 875,
|
| 380 |
"num_input_tokens_seen": 0,
|
| 381 |
"num_train_epochs": 25,
|
| 382 |
"save_steps": 500,
|
|
|
|
| 392 |
"attributes": {}
|
| 393 |
}
|
| 394 |
},
|
| 395 |
+
"total_flos": 4.376134205662298e+18,
|
| 396 |
"train_batch_size": 8,
|
| 397 |
"trial_name": null,
|
| 398 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9f044765f3dcdd375a956d71774ce12973156a4c62f7c3a5e1e5b0698ac8d8b9
|
| 3 |
+
size 5713
|